Nonexperimental designs • Description is one level of scientific inquiry – Remember • Description • Prediction • Explanation – useful for describing human behavior • but does not allow for prediction • does not allow for causal inference – valuable techniques for certain situations • When you cannot run an experimental design • Before we get into specific descriptive methods we need to discuss a couple of concepts that are important for evaluating any research methodology • Those concepts are – internal validity – external validity – quantitative approaches – qualitative approaches Internal validity • Internal validity – The degree to which a research design allows us to form causal relationships – experimental methods tend to be high – nonexperimental methods tend to be low external validity • external validity (generalizability) – Do the findings from a research project generalize to people in real world settings? • nonexperimental descriptive approaches tend to be higher – often done with humans – often done in real world settings. • Experimental/laboratory methods tend to be low – Contrived laboratory settings – Animal research • Thus, – nonexperimental methods tend to be high in external validity, but low in internal validity. • Studying children born to mothers that used cocaine during pregnancy – Experimental methods tend to be high in internal validity, but low in external validity • Studying neonatal rat pups exposed to cocaine. • research designs can also be described as primarily qualitative or primarily quantitative. – qualitative • Written descriptions. • Expressed in nonnumerical terms. – quantitative • Data is expressed numerically • The designs we will discuss in this section are nonexperimental and their goal is to describe. • Tend to be low in internal validity, but higher in external validity. • Also tend to rely on qualitative data • We will discuss four approaches – – – – phenomenology case studies field studies survey research • Phenomenology – Similar to Wundt and Titchener’s introspection. – A description of one’s own immediate experience • Not a scientific method – not publicly observable • Nevertheless has been used as a first step in developing research questions – Purkinje • Purkinje shift (1825) – noticed that the brightness of colors changed at the end of the day as the sun was going down – day • reds and yellows appeared brighter than blues and greens – twilight • blues and greens appeared brighter than reds and yellows • duplex theory of vision – cones – rods • benefits of phenomenology – can lead us to new hypotheses and discoveries – can be useful technique to ask participants about the study after they get done. • how the felt about the study – bored? – difficult? • problem with phenomenology – not public (private experience) • difficult to know if others would have the same experience – therefore any observations would have to be verified with a scientific method. • Case studies or case history • A descriptive record of an individuals experiences and/or behaviors kept by an outside observer • Clinical psychologists often use case studies • Kazdin (1992) has argued that case studies serve five major purposes – 1) They are a source of inferences, hypotheses, and theories • The dopamine hypothesis of schizophrenia was developed from medical case studies – drugs that worked for psychosis caused Parkinson’s symptoms • 2) Source for developing therapy techniques – try a new therapy and document its performance • doesn’t constitute hard evidence that it is effective • that would require control groups • 3) They allow the study of rare phenomena – Luria (1968) • the mind of a mnemonist – H. M. (surgery in 1953) • 4) They provide exceptions, or counter instances, to accepted ideas theories or practices – Gazziniga (1972) Split brain patients • we tend to believe that we have one unified consciousness • there is evidence that the two hemispheres represent separate consciousness that cooperate – corpus callosum • reach in bag • touch object with right hand – can tell you what it is – right hand goes to left brain • touch with left hand – can’t reply verbally, but can point to correct • 5) they have persuasive and motivational value – Endel Tulving (forms of memory) • Episodic, Semantic,Procedural – H. M. • Drawbacks of case studies – When it is a rare disorder it is difficult to know if others will show similar effects • small N – When we rely on self report people may leave things out that they don’t think are important or are embarrassed about. – when they rely on retrospective data (memory) information can be inaccurate • people may not remember well • memories can be altered or reconstructed – by the way questions are asked. • Field studies – naturalistic observation – systematic observation – participant observer studies – naturalistic observation • Jane Goodall – Chimps • Gans (1962) – – – – studied people living in slums watched people listened to conversations attended meetings • found people were opposed to having the old buildings replaced • Naturalistic observation requires the researcher to be as unobtrusive as possible. • It is also primarily a qualitative method • Of course watching people in natural settings leads to high rates of external validity. • But with no manipulation of variables the internal validity is low. • Systematic observation is a more quantitative approach to naturalistic observation • the researcher creates a prearranged strategy for recording observations. • The observations are recorded using specific rules • • • • Bakeman and Brownlee (1980) social behavior in children Videotaped three year olds playing together taped for 100 minutes – observers viewed the tape and recorded behavior every 15 seconds according to the following criteria • 1) unoccupied: child is not doing anything in particular or is simply watching other children • 2) solitary play: Child plays alone with toys, but is uninterested in or affected by the activities of other children • 3) together: child is with other children bus in not occupied with any particular activity • 4) parallel play: child plays beside other children with similar toys but does not play with others • 5) group play: child plays with other children, including sharing toys, or participating in organized play activities as part of a group of children. • Type a number into the computer 1-5 every 15 seconds. – now you have quantitative data – which activity occurred the most? – what order do activities occur in? Bakeman and Brownlee noted the sequence of play activities • children rarely went from being unoccupied to group play • often went from parallel to group play • perhaps parallel play is a transitional state. • participant observer studies – This is a form of naturalistic observation in which the researched pretends to be a member of the group they are studying • Jane Goodall...and chimps might be an example. – Humphreys (1970) • befriended gay men that were meeting and having sexual encounters in public restrooms – served as a lookout – but was really collecting data • even wrote down license plate numbers, obtained addresses, disguised himself, and visited their homes to interview them. – unethical • Methodological issues for all forms of field studies • Reactivity – The presence of an observer may effect behavior • unobtrusive measures – one way mirror – hidden camera – 90 degree lens --- eyebrow arch • Reliability – the degree to which a measure is stable or consistent • inter-rater reliability • two people must score the data – if the scores are very similar you have a reliable measure • requires clear operational definitions • don’t make things too complicated – my Master’s project » 6 chambers » turning, twitching, locomotion, rearing, yawning, locomotion against wall. • Survey Research – rather than simply observing you directly question your participants about their behavior, attitudes, beliefs, and intentions • past, present, or future – two major considerations • sampling technique • construction of questionnaire • population vs. sample • Surveys are concerned with external validity. – does the data reflect how the population actually feels – to ensure that it does the sample has to be representative • convenience • simple random sampling • stratified sampling – form subgroups and sample randomly from each » Freshman, sophomores, juniors, seniors • proportionate sampling – form subgroups and sample randomly, but only as many in each group as is in actual population • freshman (31), sophomore (25), junior (14), senior (19), other (11) • cluster sampling – usually used with bigger populations • randomly select courses to be sampled • everyone in course would get the survey – better example might be school districts in Indiana. – every individual in district gets surveyed if selected. • open ended vs. closed ended questions • open ended – what do you think of parking on IPFWs campus • closed ended – How difficult is it to find a parking spot on IPFWs campus • • • • a) easy – parking is readily available b) somewhat easy – some days I have to look for a while c) difficult – most days I have to look for a while d) very difficult – Every day I spend a significant amount of time looking for a spot – likert scale • On a scale of 1-7 with 1 being very easy and 7 being nearly impossible how would you rate the ease at which you can find a parking space on IPFWs campus – 1...2...3....4...5....6....7